190 THE TREND OF THE RACE 
has been going on for a long time it is supposed that the greater 
preponderance of the dolichocephalic race in the city population, 
and especially in the higher levels of wealth and culture is the 
result of the action of natural selection in favor of the superior 
type. The city draws the best of the country stock, and of the 
inhabitants that have migrated to the country the more dolicho- 
cephalic succeed best in the struggle for wealth and power. 
We may admit that Ammon has shown that in Baden changes 
have been taking place in the characteristics of the inhabitants. 
It is not so clear, however, that these changes have been chiefly 
the result of natural selection. The racial composition of com- 
munities is very apt to change as the result of migration and the 
operation of differential fecundity. Many of us have witnessed in 
this country a marked change in the character of the population 
of restricted localities within a period of a few decades. And it 
is quite evident that such changes are not due to natural selection. 
Observation of a change in the inhabitants occurring in a small 
area and in a comparatively short interval of time will not offer 
conclusive evidence regarding the factors producing the change. 
Most of the anthropometric data assembled to prove the opera- 
tion of natural selection is not convincing in that it does not 
exclude the operation of other possible causes. 
Any consideration of the réle of natural selection in man must 
take account of the much discussed question of the selective 
nature of the infant death rate. The first year is by far the most 
precarious period of life. The infant mortality rate varies enor- 
mously in different countries, according to social and economic 
conditions and the general enlightenment of the inhabitants. 
In Chile in 1903 it was over 352 per thousand births. For several 
decades in most countries of Europe the infant mortality rate has 
been somewhere between 100 and 200 per thousand. It is high in 
Prussia, Austria, Hungary and Russia, but exceptionally low in 
Norway and Sweden. It is low in Australia and lowest of all in 
New Zealand where it reached the remarkable figure of 51 in 1912. 
The infant mortality of the United States has been estimated at 
124 for 1910, although in the absence of data on the birth rate 
